[jira] [Comment Edited] (SPARK-40582) NullPointerException: Cannot invoke invalidateSerializedMapOutputStatusCache() because "shuffleStatus" is null

["Garret Wilson (Jira)" <ji...@apache.org>]

    [ https://issues.apache.org/jira/browse/SPARK-40582?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17610163#comment-17610163 ] 

Garret Wilson edited comment on SPARK-40582 at 9/27/22 5:14 PM:
----------------------------------------------------------------

{quote}Do you use Scala 2.13?{quote}

Yes. (Sorry; I misread the question earlier.)

{code:xml}
<scala.version>2.13</scala.version>
<spark.version>3.3.0</spark.version>

…
<dependency>
  <groupId>org.apache.spark</groupId>
  <artifactId>spark-core_${scala.version}</artifactId>
  <version>${spark.version}</version>
  <exclusions>
    <exclusion>
      <groupId>org.apache.logging.log4j</groupId>
      <artifactId>log4j-slf4j-impl</artifactId>
    </exclusion>
  </exclusions>
</dependency>
{code}

(Don't get me started about logging. See SPARK-40489 for that.)


was (Author: garretwilson):
{quote}Do you use Scala 2.13?{quote}

Yes. (Sorry; I misread the question earlier.)

{code:xml}
<scala.version>2.13</scala.version>
<spark.version>3.3.0</spark.version>

…
<dependency>
  <groupId>org.apache.spark</groupId>
  <artifactId>spark-core_${scala.version}</artifactId>
  <version>${spark.version}</version>
  <exclusions>
    <exclusion>
      <groupId>org.apache.logging.log4j</groupId>
      <artifactId>log4j-slf4j-impl</artifactId>
    </exclusion>
  </exclusions>
</dependency>
{code}

> NullPointerException: Cannot invoke invalidateSerializedMapOutputStatusCache() because "shuffleStatus" is null
> --------------------------------------------------------------------------------------------------------------
>
>                 Key: SPARK-40582
>                 URL: https://issues.apache.org/jira/browse/SPARK-40582
>             Project: Spark
>          Issue Type: Bug
>          Components: Spark Core
>    Affects Versions: 3.3.0
>            Reporter: Garret Wilson
>            Priority: Critical
>
> I'm running a simple little Spark 3.3.0 pipeline on Windows 10 using Java 17 and UDFs. I hardly do anything interesting, and now when I run the pipeline on only 30,000 records I'm getting this:
> {noformat}
> [ERROR] Error in removing shuffle 2
> java.lang.NullPointerException: Cannot invoke "org.apache.spark.ShuffleStatus.invalidateSerializedMapOutputStatusCache()" because "shuffleStatus" is null
>         at org.apache.spark.MapOutputTrackerMaster.$anonfun$unregisterShuffle$1(MapOutputTracker.scala:882)
>         at org.apache.spark.MapOutputTrackerMaster.$anonfun$unregisterShuffle$1$adapted(MapOutputTracker.scala:881)
>         at scala.Option.foreach(Option.scala:437)
>         at org.apache.spark.MapOutputTrackerMaster.unregisterShuffle(MapOutputTracker.scala:881)
>         at org.apache.spark.storage.BlockManagerStorageEndpoint$$anonfun$receiveAndReply$1.$anonfun$applyOrElse$3(BlockManagerStorageEndpoint.scala:59)
>         at scala.runtime.java8.JFunction0$mcZ$sp.apply(JFunction0$mcZ$sp.scala:17)
>         at org.apache.spark.storage.BlockManagerStorageEndpoint.$anonfun$doAsync$1(BlockManagerStorageEndpoint.scala:89)
>         at scala.concurrent.Future$.$anonfun$apply$1(Future.scala:678)
>         at scala.concurrent.impl.Promise$Transformation.run(Promise.scala:467)
>         at java.base/java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1136)
>         at java.base/java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:635)
>         at java.base/java.lang.Thread.run(Thread.java:833)
> {noformat}
> I searched and couldn't find any of the principal terms in the error message.
> Disconcerting that Spark is breaking at what seems to be a fundamental part of processing, and with a {{NullPointerException}} at that.
> I have already asked this question on [Stack Overflow|https://stackoverflow.com/q/73732970], and even posted a bounty, with no solutions. (The only answer so far is from someone who doesn't even use Spark and just posted links.)
> _Update:_ Now it just happened with only 1000 records. But then I reran the pipeline immediately with no changes, and it succeeded. So this {{NullPointerException}} bug is nondeterministic. Not good at all.
> _Update:_ Now it just happened with only 10 records. But then as before I reran the pipeline immediately with no changes, and it succeeded.



--
This message was sent by Atlassian Jira
(v8.20.10#820010)

---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org